Digital Media Platform serving the Fortune 50 and more, reduces time-to-insight on campaign reports from 3 weeks to 30 minutes
Boasting customers from across the Fortune 50, Infillion is a full-service media buying platform, connecting a network of publishers, unique ad form factors, and technology to better target customers and measure engagement.
Supporting a business as data-driven as Infillion is their Business Intelligence team, led by Daniel Chon, Director of Business Intelligence. “My team sits in the middle of the technical and the non-technical. Engineers rely on my team to communicate business needs, but at the end of the day, we try to make data accessible and understandable,” Daniel explained.
Daniel has worked for Infillion for nearly as long as it’s existed – 15 years, when they were known as SocialVibe. “I was assigned to monitoring the health of our ad business using Excel and SQL, which I learned on the job. Learning SQL and Python and navigating our data felt like I had acquired a power,” Daniel shared. “I love sifting through data, manipulating it, and seeing if there are any patterns I can pick out.”
Infillion is the product of multiple companies, once known as TrueX and Gimbal, and has recently acquired Anakytiks.ai, a technology that tracks customer traffic within brick and mortar stores. As a result, Infillion’s data technology is modern, comprising Redshift, BigQuery, and Tableau, but has yet to be integrated into one, cohesive stack, adding complexity and effort as Daniel’s team explored and queried data through disparate interfaces.
It was for these reasons that Active Metadata Management, and ultimately Atlan, proved compelling for Infillion. “We weren’t able to query our data unless we logged into two different systems,” Daniel shared. “And with Atlan, we’re able to understand both systems within a single interface, which has made things easier. It’s helped with queries for the TrueX business, and the cataloging has helped me understand the Gimbal business better.”
Significant Effort on Custom Reports
A crucial part of Daniel’s responsibilities are supporting Infillion’s account teams with accurate, timely reports on marketing campaign performance. Their customers, typically advertising agencies, each have unique reporting formats and data groupings, making it difficult for Infillion to automate reports.
“Our team, primarily account managers, pulled the data from standard internal reports, and then they manually reformatted the data depending on client requirements,” Daniel shared. “To give them a hand, sometimes we created a custom ad performance report for them in the desired format, so they didn’t have to manipulate anything.”
While generating these custom reports was valuable to Infillion’s account managers, the process required careful attention and time from valuable resources. “I relied on two resources for reporting, Data Engineers and Tableau. Data Engineers would help me schedule email reports, and Tableau was more self-serve,” Daniel explained.
Increasing the burden of this process, Infillion’s customers expected frequent updates on performance as campaigns were in progress, rather than when their campaign concluded. As the number of these requests increased, Daniel had no choice but to offload more of this work to Infillion’s customer-facing Account Managers, who were less savvy at manipulating data. Initially a promising way to reduce the burden on the data team, supporting customer-facing teams as they worked with data meant even more effort. “It just meant more work on my plate to help them get the data into the desired format. It would take a lot of time out of my day.”
With limited resources on the Business Intelligence and Data Engineering teams to help, changing expectations from customers, and an increasing number of requests for custom reports, Daniel and his team had to find a way to lessen the burden.
Reducing Wait Time on Reports by Three Weeks with Scheduled Queries
In Atlan Insights, a metadata-based query builder, Daniel and his team found a promising solution. “When we were in our trial with Atlan, I joined a presentation where they showed a scheduled reporting feature. It’s something I’ve been looking for, for a long time. It was a way to schedule reports without any engineering help,” Daniel shared.
Now, rather than requesting help from Data Engineering to pull the right data, and creating Tableau reports for each request himself, Daniel and his team use Atlan Insights to schedule queries at whatever cadence their customers prefer. “Whenever we get a request for a report, we’ll just set up the scheduled query in Atlan ourselves,” Daniel explained.
“It’s allowed us to fully automate reporting without any engineering resources and become self-sufficient. So it’s definitely removed a big bottleneck in our process. When I put in a request to the data engineers, it would take a week to get it into their sprint, and then it would take another week or two to actually produce it. So a three-week waiting period is just not scalable. So this definitely cut that down to a half an hour, maybe less, to set up.”
Daniel Chon, Director of Business Intelligence
Beyond the clear benefits that a 3-week time savings represents to a resource-constrained data team, the higher level of service being provided to Infillion’s Account Managers, and ultimately their customers, does not go unnoticed. “They do love that we have this toolset now. Since the first scheduled Atlan report, the number of requests for our custom scheduled reports have just increased. I think it’s because users are realizing this is now part of our standard BI toolset,” Daniel shared.
Adding Context and Transparency with Atlan
Having raised the bar on how the Business Intelligence Team can support Infillion’s business, Daniel and his team are turning their focus to democratizing access to data, improving context on data assets, and further stitching together disparate systems.
“What’s next for our team is to use the feature Atlan was built for, which is the cataloging. We want to add more descriptions to tables and metrics, especially since a lot of this information is spread across many different systems,” Daniel explained. “Having this would definitely help us dive into our data much more quickly, and be a lot more self-reliant when it comes to figuring out where data lives and what it means.”
Header Photo: Product School on Unsplash
